pub struct MetaLearningEngine {
pub regret: RegretTracker,
pub plateau: PlateauDetector,
pub pareto: ParetoFront,
pub curiosity: CuriosityBonus,
pub decaying_betas: HashMap<(ContextBucket, ArmId), DecayingBeta>,
/* private fields */
}Expand description
Unified meta-learning engine that composes all five improvements.
Drop-in enhancement for the existing DomainExpansionEngine.
Call record_decision after each arm selection and check_plateau
periodically to get adaptive strategy recommendations.
Fields§
§regret: RegretTracker§plateau: PlateauDetector§pareto: ParetoFront§curiosity: CuriosityBonus§decaying_betas: HashMap<(ContextBucket, ArmId), DecayingBeta>Per-bucket decaying beta distributions (optional overlay).
Implementations§
Source§impl MetaLearningEngine
impl MetaLearningEngine
Sourcepub fn with_config(
regret_snapshot_interval: u64,
plateau_window: usize,
plateau_threshold: f32,
exploration_coeff: f32,
decay_factor: f32,
) -> Self
pub fn with_config( regret_snapshot_interval: u64, plateau_window: usize, plateau_threshold: f32, exploration_coeff: f32, decay_factor: f32, ) -> Self
Create with custom parameters.
Sourcepub fn record_decision(
&mut self,
bucket: &ContextBucket,
arm: &ArmId,
reward: f32,
)
pub fn record_decision( &mut self, bucket: &ContextBucket, arm: &ArmId, reward: f32, )
Record a decision outcome. Call after every arm selection.
Sourcepub fn record_kernel(
&mut self,
kernel_id: &str,
accuracy: f32,
cost: f32,
robustness: f32,
generation: u32,
)
pub fn record_kernel( &mut self, kernel_id: &str, accuracy: f32, cost: f32, robustness: f32, generation: u32, )
Record a population kernel’s multi-objective performance.
Sourcepub fn check_plateau(&mut self, points: &[CostCurvePoint]) -> PlateauAction
pub fn check_plateau(&mut self, points: &[CostCurvePoint]) -> PlateauAction
Check the cost curve for plateau and recommend action.
Sourcepub fn boosted_score(
&self,
bucket: &ContextBucket,
arm: &ArmId,
thompson_sample: f32,
) -> f32
pub fn boosted_score( &self, bucket: &ContextBucket, arm: &ArmId, thompson_sample: f32, ) -> f32
Get the curiosity-boosted score for an arm.
Combines the Thompson Sampling estimate with an exploration bonus.
Sourcepub fn decaying_mean(&self, bucket: &ContextBucket, arm: &ArmId) -> Option<f32>
pub fn decaying_mean(&self, bucket: &ContextBucket, arm: &ArmId) -> Option<f32>
Get the decaying beta mean for a bucket/arm (if tracked).
Sourcepub fn health_check(&self) -> MetaLearningHealth
pub fn health_check(&self) -> MetaLearningHealth
Comprehensive health check of the learning system.
Trait Implementations§
Source§impl Clone for MetaLearningEngine
impl Clone for MetaLearningEngine
Source§fn clone(&self) -> MetaLearningEngine
fn clone(&self) -> MetaLearningEngine
Returns a duplicate of the value. Read more
1.0.0 · Source§fn clone_from(&mut self, source: &Self)
fn clone_from(&mut self, source: &Self)
Performs copy-assignment from
source. Read moreSource§impl Debug for MetaLearningEngine
impl Debug for MetaLearningEngine
Source§impl Default for MetaLearningEngine
impl Default for MetaLearningEngine
Source§impl<'de> Deserialize<'de> for MetaLearningEngine
impl<'de> Deserialize<'de> for MetaLearningEngine
Source§fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>where
__D: Deserializer<'de>,
fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>where
__D: Deserializer<'de>,
Deserialize this value from the given Serde deserializer. Read more
Auto Trait Implementations§
impl Freeze for MetaLearningEngine
impl RefUnwindSafe for MetaLearningEngine
impl Send for MetaLearningEngine
impl Sync for MetaLearningEngine
impl Unpin for MetaLearningEngine
impl UnsafeUnpin for MetaLearningEngine
impl UnwindSafe for MetaLearningEngine
Blanket Implementations§
Source§impl<T> BorrowMut<T> for Twhere
T: ?Sized,
impl<T> BorrowMut<T> for Twhere
T: ?Sized,
Source§fn borrow_mut(&mut self) -> &mut T
fn borrow_mut(&mut self) -> &mut T
Mutably borrows from an owned value. Read more